MTN conducted a human rights risk assessment on Zigi, MTN Nigeria’s AI-enabled customer service chatbot, as part of its broader Responsible AI and digital human rights governance approach, guided at a high level by MTN’s BRAIN principles. The assessment was grounded in the UN Guiding Principles on Business and Human Rights and formed part of MTN’s standard AI risk management processes. Zigi supports customers with everyday interactions such as checking available offers, viewing balances, purchasing bundles and airtime, and connecting with a human agent when required. It is used for both transactional and non-transactional purposes and provides a convenient, remote alternative to in-store or call centre engagement, helping to reduce congestion in physical channels while improving service efficiency. The assessment focused on how Zigi could affect the rights to freedom of expression, access to information, privacy, consumer protection, and language and cultural identity. 

What we did 

MTN assessed both the potential benefits and risks of Zigi from a human rights perspective. The review recognised that digital chat solutions can expand access to information, reduce language barriers and enable more convenient access to services. It also considered risks such as the potential spread of incorrect information, biased or unfair outcomes in automated interactions, privacy and data protection concerns, and exposure to inappropriate content if safeguards were insufficient. To manage these risks, MTN strengthened controls around Zigi’s operation. The chatbot is being progressively enhanced to better reflect local languages, expressions and communication styles to ensure more natural and culturally appropriate engagement for Nigerian customers. All content prompts and responses are reviewed daily by human teams to ensure accuracy, reliability and alignment with MTN standards. Customer data processed through Zigi is anonymised and pseudonymised, with identification occurring only when customers log in through an authorised channel and limited to that session. Where Zigi cannot resolve a query, it is transferred to a human agent for appropriate support.

Outcomes and impact 

The assessment improved MTN’s understanding of how AI-enabled customer service tools can affect people’s rights and informed the design of stronger safeguards. Zigi now operates as a complementary service channel alongside email and call centre support, enabling more efficient customer service while maintaining strong protections for customer rights, privacy and trust. The process also reinforced cross functional co-ordination between Digital, Legal, Risk, Sustainability and Corporate Affairs teams on AI-related decision making. 

Lessons learnt 

Human rights considerations need to be integrated early in the design and deployment of AI-enabled customer tools. Regular human oversight is essential to maintaining accuracy, trust and accountability. Localisation and cultural relevance are critical for inclusive digital services in linguistically diverse markets, and ongoing monitoring and iterative improvement are necessary to ensure AI-powered solutions remain aligned with MTN’s values and digital human rights commitments